Congress on Technical Advancement 2017 127

Slope Stabilization for Local Government Engineers in Minnesota

Mitchell Nelson, S.M.ASCE1; David Saftner, A.M.ASCE2;

and Carlos Carranza-Torres, P.E.3
1
Graduate Research Assistant, Univ. of Minnesota Duluth, Duluth, MN 55812.
E-mail: nels7745@d.umn.edu
Downloaded from ascelibrary.org by Griffith University on 09/08/17. Copyright ASCE. For personal use only; all rights reserved.

2
Assistant Professor, Univ. of Minnesota Duluth, Duluth, MN 55812. E-mail:
dsaftner@d.umn.edu
3
Associate Professor, Univ. of Minnesota Duluth, Duluth, MN 55812. E-mail:
carranza@d.umn.edu

Abstract

The purpose of this paper is to describe development of a design guide focusing on

stabilization of locally-maintained slopes. Motivation for this project came from
engineers experiencing recurring slope failures, especially following heavy rain
events or spring thawing. There is currently no guide for public works engineers to
stabilize slopes of the scale typically seen along locally maintained roadways in
Minnesota. Therefore, slope failures can damage roads, pose safety hazards, and
introduce preventable maintenance and repair costs. While there is no single
stabilization method appropriate for all situations, several methods have proven
effective. Local government engineers requested guidance on stabilization methods
available without paying for specialty equipment or hiring a geotechnical consultant.
This study addresses the need to provide a consistent, logical approach to slope
stabilization founded in geotechnical research and applicable to common slope
failures. First, the authors identified slopes for further analysis via a survey sent to
each county engineering department in Minnesota and conducted site investigations to
characterize sites experiencing recurring failures. Laboratory and in situ testing
quantified slope geometry and soil strength properties for each of the slopes requiring
maintenance. Additionally, limit equilibrium method models were developed for each
slope to investigate different stabilization methods in a parametric study. Finally,
modeling and analysis results were summarized in a design guide for distribution to
local government engineers. The deliverable will assist with efficient stabilization
method selection for common slope failures along roadways.

INTRODUCTION

Currently, there is no guide for public works engineers to stabilize slopes of the scale
typically seen along locally maintained roadways. Therefore, slope failures can block
roads, pose safety hazards, and introduce preventable maintenance costs. While no
single stabilization method is appropriate for every situation, several methods have
proven effective. Researchers used site visits and characterization, laboratory testing,
and modeling simulations to produce a slope stabilization guide for local government

© ASCE

Congress on Technical Advancement 2017

 Congress on Technical Advancement 2017 128

engineers. The deliverable recommends simple, effective methods of stabilizing

slopes and repair options for common, recurring slope failures.

A survey of local government engineers in Minnesota provided the opportunity to

develop case study sites. Soil characterization of samples from each site provided
representative strength parameters for limit equilibrium method (LEM) modeling.
Analysis of models with different stabilization methods implemented in a parametric
study led to the recommendations in the project deliverable.
Downloaded from ascelibrary.org by Griffith University on 09/08/17. Copyright ASCE. For personal use only; all rights reserved.

BACKGROUND

To provide slope stabilization recommendations, researchers needed to establish a

background understanding of slope stability and stabilization methods. Authors
performed a literature review to identify stabilization approaches for research.

Slope Stability Overview

Slope stability is typically quantified with a factor of safety (FS). Stabilizing a slope
involves increasing the FS. The FS is the ratio of shear strength to the required shear
strength for equilibrium along a given potential failure surface. Fundamentally,
increasing the FS involves introducing more stabilizing forces (increasing capacity) or
limiting driving forces (decreasing demand). Academic research and standard
engineering practice have identified many slope stabilization methods; most fit into
four categories: controlling groundwater with drainage, using surface cover,
excavating and regrading, and adding reinforcing support structures. A literature
review was conducted to determine effective ways to stabilize the slopes maintained
by public works engineers.

Controlling Water
Water limits soil’s ability to resist shearing, which can lead to slope failure. An
increase in pore pressure causes a decrease in effective stress (σ’). Because σ’
governs the soil’s strength characteristics, the presence of water leads to decreased
soil shear strength. Drainage can minimize the amount of water present in the slope.
Drains provide a path for water to flow away from the potential slide area and
increase shear strength. Surface drains, trenches, horizontal drains, and drainage
wells are methods to control water in the slope area (Cornforth, 2005). Site
conditions and contractor experience often determine dimensions, layout, and spacing
of drainage features. Nelson (2017) details drainage features and stabilization
methods to remove groundwater.

Surface Cover
Another stabilization method is surface cover. Appropriate soil cover can divert
water, limit the effects of erosion, and provide stabilizing forces for the upper layer of
a slope. Vegetative cover, rip-rap, and buttressing are common approaches to slope
stabilization by ground cover. Grass and other vegetation protect the soil from the
impact of rainwater and surface runoff. Operstein et al. (2000) describes how
vegetation affects soil shear strength, and describes the mechanical stabilization that

© ASCE

Congress on Technical Advancement 2017

 Congress on Technical Advancement 2017 129

roots provide. Coarse rock rip rap is also a material option for surface cover.
However, rip rap cover methods can add weight and actually decrease global stability;
rip rap cover is primarily an erosion-control method. Buttressing is placing a soil or
rock mass against a slope face to add stabilizing force and decrease the overall slope
height, as Figure 1 shows.
Downloaded from ascelibrary.org by Griffith University on 09/08/17. Copyright ASCE. For personal use only; all rights reserved.

Figure 1. Buttress example

Decreasing Load / Regrading

If spatial concerns such as jobsite and right-of-way boundaries are not an obstacle,
decreasing the slope inclination angle can increase stability. Also, using lightweight
fill can decrease slope weight and lower driving forces. Abramson et al. (2002)
identifies expanded shale, shredded tires, encapsulated sawdust, seashells, and
polystyrene foam as some examples. Replacing slope material with free-draining
sand fill minimizes uncertainty in ground conditions and improves drainage.

Reinforcing Support Structures

Installing reinforcing structures generally increases FS. Retaining walls, soil nailing,
ground anchors and mechanically stabilized earth (MSE) walls are examples of
stabilizing structures (Abramson et al., 2002). Most reinforcing structures require
specialized experience and are likely more suitable solutions for large projects.
Although reinforcing structures are an expensive option, they are sometimes
necessary to stabilize slopes.

METHODS

County Engineer Survey

Ensuring that recommendations were practical for local government agencies was
critical. Accordingly, researchers developed a survey for Minnesota county engineers
to determine common stabilization methods, identify slopes requiring maintenance,
and schedule site visits to develop case studies (Nelson, 2017). Minnesota county and
maintenance engineers received the survey in September 2015. Fourteen engineering
departments responded, indicating a variety of stabilization methods used.

Site Investigation and in situ Testing

Figure 2 summarizes the site investigation locations. During site visits, researchers
measured slope geometry with a field tape measure and surveying equipment and
determined soil type and strength properties. Vane shear test and pocket

© ASCE

Congress on Technical Advancement 2017

 Congress on Technical Advancement 2017 130

penetrometer results indicated the value of undrained shear strength, and soil was
visually classified. Researchers collected in situ measurements of soil strength by
conducting the dynamic cone penetrometer (DCP) in general accordance with ASTM
D7380 – 15. The DCP test results are comparable to standard penetration test (SPT)
results and the test is more easily implemented than full-scale geotechnical field
testing.
Downloaded from ascelibrary.org by Griffith University on 09/08/17. Copyright ASCE. For personal use only; all rights reserved.

Figure 2. Locations of case study site investigations

Ojakangas (2009) and other sources provided geologic background for each site.
Depth to the ground water table was estimated using the Minnesota Department of
Natural Resources (DNR) monitoring wells. Researchers also classified observed
slope failures. Most commonly observed failures were creep failures and rotational
slide failures. The goal of the site investigations was to compile a representative set
of case studies for analysis and modeling.

Laboratory testing
Researchers determined soil strength properties using the direct shear test in
accordance with ASTM D3080-11. Slope failures are examples of plane strain, and
direct shear test specimens exhibit the same failure mechanism. This similarity makes
slope stability failure modeling a good application of direct shear testing. The
outcome of direct shear testing was values of shear strength parameters for each soil
sample, particularly effective friction angle (φ’) and effective cohesion (c’).
Researchers conducted Atterberg Limit testing in general accordance with ASTM
D4318-10 to determine the plastic limit (PL) and liquid limit (LL) of each sample.
ASTM 2487-11 describes how to use these parameters to classify fine samples. With

© ASCE

Congress on Technical Advancement 2017

 Congress on Technical Advancement 2017 131

sieve analysis used to classify soils containing granular soil, the research team
assigned Unified Soil Classification System (USCS) classifications to each sample.

Slope Modeling
Researchers performed slope stability modeling using the Rocscience program
SLIDE. The program determines slope FS using the method of slices. The output
from each model is a rendering of the slope and site conditions, the lowest computed
FS, and the critical failure surface, as Figure 3 shows. By comparing the baseline FS
to the FS of the same site with a stabilization method implemented, researchers could
Downloaded from ascelibrary.org by Griffith University on 09/08/17. Copyright ASCE. For personal use only; all rights reserved.

quantify the effectiveness of the method. Modeling the slope with a different
stabilization technique and using the same quantitative analysis allowed researchers to
determine the most effective method. Following this parametric study approach,
researchers investigated scenarios for which each technique was most effective;
noting common input conditions and examining FS output from each led to
recommendations used to develop the project deliverable.

Figure 3. Example SLIDE output

The output from LEM modeling is the slope’s FS against global rotational failure.
Therefore, at sites exhibiting creep failure, the modeling output is not always an
accurate simulation of site characteristics. Infinite slope analysis uses a more simple
calculation that considers slope inclination angle (β), φ’ and soil unit weight (γ). The
equation for infinite slope FS for dry slopes is shown in Equation 1, and FS for
saturated slopes is shown in Equation 2.
tan(φ’) γ − γ tan(φ’)
(Eqn. 1) = (Eqn. 2) =
tan(β) γ tan(β)
Since no samples were clean sand, all slope material observed had some cohesion;
therefore, the FS from infinite slope analysis was conservative. Some SLIDE outputs
showed failure surfaces with large failure radii, indicating that infinite slope analysis
is a better way to asses FS.

RESULTS & DISCUSSION

Case Study Sites and Field Investigation Summary

Researchers investigated and documented fourteen sites in Minnesota. Nelson (2017)
describes each site investigation and the unique aspects of each site. Of the

© ASCE

Congress on Technical Advancement 2017

 Congress on Technical Advancement 2017 132

documented sites, researchers noted five with primarily sandy soil, eight with
primarily fine-grained soil, and one rock site. Slope failure was observed at nine
sites, while four sites were already stabilized. The damaging effects of groundwater
were visible in most site failures, indicating that controlling water is a valuable
stabilization method. One site bridged a stream with a culvert that appeared to fail
and cause slope damage, and three sites showed slope stability issues in back slopes.

Soil Strength Quantification

Laboratory analysis provided a comprehensive background of materials at case study
Downloaded from ascelibrary.org by Griffith University on 09/08/17. Copyright ASCE. For personal use only; all rights reserved.

sites. Values of c’ and φ' that were critical to modeling came from direct shear testing
on samples collected during field investigations. Due to extreme slope geometry and
severe slope failure, the team excluded the sites in St. Louis and Washington Counties
from lab and modeling analysis. Geotechnical consulting and additional
considerations would be necessary to stabilize these slopes, and researchers limited
the scope of analysis to common, recurring slope failures.

Slope Stability Modeling

Researchers produced LEM models of each site with various stabilization methods in
place for comparison. The literature review identified several stabilization methods
that increase slope stability including retaining walls, soil nailing, geosynthetics and
MSE walls. These techniques are effective when applied correctly. In some cases it
is appropriate to consider such methods. These solutions, however, are outside the
scope of research; the guide aims to make recommendations for simple in-house
stabilization options that do not require hiring a geotechnical consultant or conducting
specialty analysis.

Validating Models and Infinite Slope Calculations

Modeling results were used to conduct a parametric study, not necessarily to describe
an actual site’s stability. When modeling sites with observed failure, researchers
expected the output FS to be less than or equal to 1.0, confirming failure. LEM
simulations of some failed slope situations (where FS should be less than 1.0) resulted
in FS values greater than 1.0, indicating that for some scenarios, especially creep
failure sites, LEM modeling can over-estimate the FS. Poor compaction, freeze-thaw
cycling, and undocumented fill all lower soil strength, but are not directly addressed
in SLIDE input parameters. Therefore, for sites with observed failures, researchers
used decreased strength values to model conditions noted in the field.

Sites with creep failure were more difficult to validate because the output from
SLIDE identifies the circular plane with the lowest resistance to sliding. Infinite
slope analysis was performed to represent sites exhibiting creep failure. Results of
infinite slope analysis were considered where more applicable than LEM outputs. For
these sites, the infinite slope FS was used as a baseline for comparing stabilization
methods. The results of infinite slope analysis for appropriate sites are presented by
Nelson (2017). Authors considered both analysis types, but only used one for each
site when determining the baseline for parametric analysis.

© ASCE

Congress on Technical Advancement 2017

 Congress on Technical Advancement 2017 133

Modeling Groundwater and Drainage Features

Researchers modeled the effects of groundwater by considering a steady-state, worst-
case scenario. Assuming all failed drainage features or soil with poor drainage
properties, the groundwater table would rise to the ground surface. Figure 4
demonstrates modeling the scenario.
Downloaded from ascelibrary.org by Griffith University on 09/08/17. Copyright ASCE. For personal use only; all rights reserved.

Figure 4. Modeling groundwater effects; a) hydrostatic and b) worst-case water

table condition
The example demonstrates the difference between hydrostatic conditions (a) and poor
drainage (b) in terms of output FS. Researchers did not consider transient
groundwater analysis to represent a situation somewhere between the two conditions.
Any construction below the water table depth should involve detailed geotechnical
analysis; such scenarios are outside the scope of research.

Modeling Surface Cover

Stabilizing the uppermost soil layer minimizes the effect of soil creep and limits
pavement damage. For modeling, the research team considered replacing the top foot
of in situ soil with fill material. Researchers executed models with properties for
course gravel and cobble rip rap. Representative strength properties for common rock
rip rap are (Attia et al., 2009): φ’ = 45°, c’ = 5 psf, γ = 120 pcf (see Figure 5).
Surface cover does not typically increase the FS; in the case of rip rap cover, the
method can increase weight and driving forces, which decreases the FS. Erosion
protection is the main benefit, which is difficult to quantify and comes at the cost of
increasing forces driving failure.

© ASCE

Congress on Technical Advancement 2017

 Congress on Technical Advancement 2017 134
Downloaded from ascelibrary.org by Griffith University on 09/08/17. Copyright ASCE. For personal use only; all rights reserved.

Figure 5. Modeling surface cover; a) without and b) with rip rap cover

Vegetative cover is another stabilization method. Operstein et al. (2000) concluded

that soil with vegetation has higher shear strength and that plant roots affect strength
by increasing overall effective cohesion. Observations from LEM modeling indicate
that c’ governs the depth of the failure surface; a soil with higher c’ will have a deeper
circular failure. Because plant roots have a quantifiable impact on c’, researchers
recommend adding vegetative cover to slopes to increase surficial stability.

Modeling Buttressing
The advantage of a buttress is that no excavation or slope reconstruction is necessary.
The same common borrow rip rap considered for surface cover can be used for
buttress material, with the same material properties. Figure 6 shows an example of
modeling buttressing. As the example shows, an aggregate buttress affects the failure
surface. Because buttress material has higher strength properties the failure occurs in
the soil. Researchers noted the most benefit in small slopes.

Figure 6. Modeling buttressing; a) without and b) with buttress

© ASCE

Congress on Technical Advancement 2017

 Congress on Technical Advancement 2017 135

Modeling Regrading
Changing slope geometry, particularly decreasing slope angles, can reduce driving
forces. Regrading, even when not changing the overall slope angle, can increase the
overall FS. The standard practice of re-compacting surface soil in benches, then
finishing the slope to a specified grade generally adds stability. Regrading is a way of
‘smoothing out’ irregularities (Figure 7). For every method that required excavation,
researchers assumed the benefit of regrading, proper construction, and re-compacting.
When comparing ‘before’ and ‘after’ models of failed slopes, researchers were able to
model soil with higher strength parameters after regrading and re-compacting.
Downloaded from ascelibrary.org by Griffith University on 09/08/17. Copyright ASCE. For personal use only; all rights reserved.

Figure 7. Modeling regrading; a) before and b) after regrading and re-

compacting

Modeling Soil Replacement

Clean, free-draining sand has ideal properties for roadway embankments. Authors
conducted direct shear testing on coarse, compacted sand to consider as replacement
material with the following properties: φ’ = 35°, c’ = 100 psf, γ = 120 pcf. The
remove-and-replace method requires excavation, but typically not a specialty
contractor. An important benefit of sand fill is improving drainage properties. For
modeling, researchers considered worst-case drainage and adequate in situ drainage
conditions for each site. The worst-case drainage situations were simulated by
placing the water table immediately at the bottom of the fill layer, assuming the native
material had poor or no drainage capability. The adequate in situ drainage situations
were executed with the water table at its baseline depth. This represents a native
material with good drainage properties, or teams installing drainage features in
addition to sand fill. The example below shows the two drainage scenarios for
replacing the top 10 feet of the same slope with sand. Figure 8 shows the worst-case
drainage scenario and Figure 9 shows the model assuming adequate drainage for the
same site.

© ASCE

Congress on Technical Advancement 2017

 Congress on Technical Advancement 2017 136
Downloaded from ascelibrary.org by Griffith University on 09/08/17. Copyright ASCE. For personal use only; all rights reserved.

Figure 8. Modeling sand replacement, worst-case drainage scenario; a) before

and b) after replacement

Figure 9. Modeling sand replacement, adequate drainage scenario; a) before and

b) after replacement

The worst-case scenario in Figure 8 simulated replacing the top portion of the slope
with free-draining fill where native material had poor drainage. The adequate
drainage scenario shown in Figure 9 represents choosing to install drainage features;
this will have a higher cost, but appears to be much more effective when in situ
drainage is poor. Sand fill should be covered after regrading to prevent erosion.

Analysis, Observations, and Development of Deliverable

The intent of modeling was to investigate slope failure and mitigation techniques to
suggest approaches for common slope failure situations. The guide shows common
slope failure types and site conditions. Users will find the set of conditions that most
closely match observed slopes, and the guide will provide recommendations based on
project results. The tool uses distinctions in three site conditions to characterize any
given slope project: failure type (i.e. soil creep or rotational failure), soil type (i.e.
cohesive or granular soil), and drainage condition (i.e. presence or absence of

© ASCE

Congress on Technical Advancement 2017

 Congress on Technical Advancement 2017 137

groundwater indicator). These distinctions most clearly categorize site conditions that
researchers observed during site investigations. Combining the site condition
distinctions yields eight scenarios. The guide features a flowchart for users to
determine which scenario to consider (Figure 10)
Downloaded from ascelibrary.org by Griffith University on 09/08/17. Copyright ASCE. For personal use only; all rights reserved.

Figure 10. Flowchart outlining the slope stabilization guide

Authors expect end users to compare any given slope stabilization site to the
scenarios provided in the guide layout. Researchers’ parametric study led to the
recommendations in the deliverable. The guide, along with background information
on analysis and scenario descriptions, is available in Nelson (2017).
CONCLUSIONS

This research project addresses needs for slope stabilization recommendations

identified by local government engineers in Minnesota. Authors used input from the
target audience, standard engineering research practices, and comparative analysis to
help engineers with slope stabilization issues. First, the research team identified case
histories representative of the project scope via a survey of Minnesota county
engineering departments. Next, authors researched various stabilization methods in a
literature review. Additionally, laboratory testing characterized soil properties and
provided strength parameters. LEM Models were developed and used to investigate
the effect of various slope stabilization methods. A parametric study of each

© ASCE

Congress on Technical Advancement 2017

 Congress on Technical Advancement 2017 138

stabilization method and each site model led to stabilization recommendations.

Authors expect the stabilization guide to assist local government engineers in
effectively using budget and time resources.

ACKNOWLEDGMENTS

Support for this research came from MnDOT Contract No. 99008, Work Order No.
190. Minnesota DOT and county engineers serving on the Technical Advisor Panel
provided insight and guidance throughout the project, particularly Blake Nelson.
Downloaded from ascelibrary.org by Griffith University on 09/08/17. Copyright ASCE. For personal use only; all rights reserved.

Researchers gratefully acknowledge the survey respondents that provided case study
sites and representatives that met with the team.

REFERENCES

Abramson, L. W., Lee, T. Sharma, S., Boyce, G. (2002). Slope stability and
stabilization methods (2nd ed.). New York, NY. Wiley.
ASTM D2216-10. (2010). Standard Test Methods for Laboratory Determination of
Water (Moisture) Content of Soil and Rock by Mass. West Conshohocken, PA.
ASTM International.
ASTM D2487-11. (2011). Standard Practice for Classification of Soils for
Engineering Purposes (Unified Soil Classification System). West Conshohocken,
PA. ASTM International.
ASTM D3080-11. (2011). Standard Test Method for Direct Shear Test of Soils
Under Consolidated Drained Conditions. West Conshohocken, PA. ASTM
International.
ASTM D4318-10. (2010). Standard Test Methods for Liquid Limit, Plastic Limit, and
Plasticity Index of Soils. West Conshohocken, PA. ASTM International.
ASTM D7380-15. (2015). Standard Test Method for Soil Compaction Determination
at Shallow Depths Using 5-lb (2.3 kg) Dynamic Cone Penetrometer. West
Conshohocken, PA. ASTM International.
Attia, M., Abdelrahman, M., Alam, T. (2009). Investigation of Stripping in Minnesota
Class 7 (RAP) and Full Depth Reclamation Base Materials. Fargo, ND. Dept.
of Civil Engineering, North Dakota State University.
Cornforth, D. (2005). Landslides in practice: Investigations, Analysis, and
Remedial/Preventative Options in Soils. Hoboken, NJ. John Wiley and Sons.
Nelson, M. (2017). Slope Stabilization for Local Government Engineers in
Minnesota. (Master’s Thesis, University of Minnesota). Duluth, MN.
Ojakangas, R. (2009). Roadside Geology of Minnesota. Missoula, MT. Mountain
Press Publishing Company.
Operstein, V. & Frydman, S. (2000). The influence of Vegetation on Soil Strength.
Ground Improvement, 4(2), 81-89.
Rocscience. (2016). SLIDE Version 7. Toronto, ON

© ASCE

Congress on Technical Advancement 2017